The previous studies showed that mislocalisation of a brief flash stimulus presented temporally proximate to the saccadic eye movement may occur because of the mismatch of the extraretinal eye position signal and the object image displacement on the retina (Honda, 1989, Perception & Psychophysics, 45: 162-174). However, the saccadic localisation was more accurate in the presence of the exocentric cue, suggesting that localisation is based on a combination of exocentric and egocentric cues (Dassonville et al., 1995, Vision Research, 35: 2191-2199). Furthermore, the final phantom flash of the repeatedly flashing stimulus (500 Hz), which was extinguished at various points of time after saccade onset, was observed near the attended positions during the saccade (Ebisawa & Fujiwara, submitted). This suggested that mislocalisation of the flashing stimulus is a function of the attended position rather than the efference and afference copies related to the saccade. The purposes of this study are to investigate the influence of the existence of the fixation point and saccade target as exocentric cues and to investigate whether or not the attended position influences localisation of a single-flash stimulus.

Two experiments were conducted with subjects' heads immobilised: "Dark" and "Background" experiments. Each experiment consisted of "Predictable" and "Unpredictable" conditions. In the Background experiment, the fixation point (FP) and saccade target (ST), were presented consistently near the time of the saccade. The subject made a saccade from FP (0 deg) to ST (8 deg). In the Dark condition, the two stimuli were extinguished before the saccade, which was made from the extinguished FP position to the remembered ST position. In both experiments, each of the four stimuli located -4, 0, 4, and 12 deg, respectively, flashed for 1 ms at various points of time near the saccade (approximately from 200 ms before until 200 ms after saccade onset). In the Unpredictable condition, the subject did not know either the time or position of the flash stimulus, whereas in the Predictable condition, the positional information of the flash stimulus was given to the subject visually at the beginning of each trial.

The results of three subjects commonly showed that in both Background and Dark experiments, the degree and duration of mislocalisation decreased in the Predictable condition compared to the Unpredictable condition. In addition, for both Predicable and Unpredictable conditions the degree and duration of mislocalisation in the Background experiment decreased in comparison to those of the Dark experiment. The difference in the mislocalisation between the Background and Dark conditions suggested that the visible stimuli (FP and ST) work as the positional references, supporting the suggestion of Dassonville et al. (1995). The difference between the Predictable and Unpredictable conditions suggested that the attention shifts influence the mislocalisation concerning a single-flash stimulus as well as the repeatedly flashing stimulus (Ebisawa and Fujiwara, submitted). This is because the subjects probably directed their attention to each known flash position in order to accurately determine the apparent position of the flash in the Predictable condition.